One day after Citadel’s Ken Griffin told investors they are “almost certainly” in an AI bubble, Deutsche Bank CEO Christian Sewing offered a different read. He doesn’t see one.
Two days, two very different takes
Griffin went first. On October 20, 2026, speaking at the Citadel Future of Global Markets Conference, the billionaire hedge fund founder warned that markets were likely in bubble territory.
His argument centered on a gap between spending and payoff. Corporations are pouring money into AI tools from providers like Anthropic and OpenAI, he said, but many are struggling to turn that investment into real productivity gains.
Griffin also drew on history. He pointed to the pattern of major economic transformations producing bubbles along the way, with the dot-com boom as the obvious reference.
Sewing responded on October 21, 2026, arguing that fears of an AI bubble are unjustified. In his view, the market volatility and rising interest rates rattling investors have more to do with geopolitical tensions than with anything specific to AI.
What Deutsche Bank’s research actually says
On the cautious side, the bank’s analysts acknowledge that some private AI company valuations look stretched. They also note that the capital expenditures required to build out AI are substantial.
On the supportive side, the research argues that financial metrics back up the public valuations of industry leaders like Nvidia and Microsoft. That, the analysts say, sets the current moment apart from the dot-com peak of the late 1990s, when valuations ran far ahead of fundamentals.
Still, Deutsche Bank’s analysts flag data-center economics as a key risk, along with the possibility that revenue fails to keep pace with spending.
Their research highlighted potential revenue shortfalls of $800 billion relative to AI-related capital expenditures. The analysts frame this as a systemic concern that could evolve through 2027, with potential knock-on effects for broader measures like GDP growth.
What this means for investors
If Griffin is right, the risk lies in the corporate customers. Should businesses fail to see returns on their AI spending, they may eventually pull back. That would ripple upstream to the companies selling the chips, cloud capacity, and models.
The signal worth tracking is the one both camps agree on: whether AI revenue starts closing the gap with AI spending. Deutsche Bank’s potential $800 billion shortfall figure gives investors a rough sense of how wide that gap could be.
Deutsche Bank separates stretched private valuations from better-supported public ones, and that line could matter if sentiment sours. Not every AI-linked asset carries the same risk, even if they tend to move together on headlines.
Meanwhile, Sewing’s point about geopolitics shouldn’t get lost in the bubble debate. If he’s correct that tensions abroad are driving rate moves and volatility, then investors focused solely on AI valuations may be watching the wrong scoreboard.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.